Backer David, Billing Trey
Center for International Development and Conflict Management, University of Maryland, United States.
Department of Government & Politics, University of Maryland, United States.
Data Brief. 2021 Mar 26;36:106999. doi: 10.1016/j.dib.2021.106999. eCollection 2021 Jun.
The Famine Early Warning Systems Network (FEWS NET) has been appraising food security in numerous countries around the world since 1985. Multiple times per year, FEWS NET reports scores for current situation assessments and future projections of food security. The scores are measured on a five-level index scale and gauged for the geographic units of livelihood zones. These zones vary in size and do not remain static, which complicates comparison of food security within and across countries and over time. To facilitate such analysis and interoperability with other sources, we transformed available raw data to the units of geospatial grid-cells that have a uniform, static resolution of 0.5° × 0.5°, a common format of data used in research across diverse disciplines. FEWS NET provides public online access to shapefiles reflecting reports back to 2009. Separate shapefiles capture assessments and projections, with further delineation by the index score. Each shapefile can comprise a complex (multi)polygon, without clear differentiation among livelihood zones. Overlaying a geospatial grid allows disaggregation of the (multi)polygons to standard units. We performed the transformation to grid-cells on the shapefiles for all 25 countries (including Yemen) that FEWS NET tracked within regional groupings of East, Southern, and West Africa from July 2009-October 2020. For each report cycle, each grid-cell was assigned scores of the assessment and near-term and medium-term projections, based on the raw data for the corresponding livelihood zone. In addition, we calculated a value of bias in medium-term projections relative to subsequent assessments, which can be used as a metric for validation of accuracy. This article provides access to the grid-cell data on assessment and projection scores and bias values. In addition, we present time-lapse animated maps as tools to visualize historical patterns and trends in these indicators across Africa. Our related research article employed the grid-cell data to evaluate the accuracy of FEWS NET projections, including as a function of variation in humanitarian assistance, climate conditions, and violent conflict (Backer and Billing [1]). Researchers can likewise use the grid-cell data to conduct further validation of food security projections and to examine the relationship of assessments and projections to potential drivers and consequences. The data and animations are also valuable to stakeholders throughout the international community seeking to learn and disseminate knowledge about the tendencies of food security projections on a broad scale.
自1985年以来，饥荒早期预警系统网络（FEWS NET）一直在评估世界上许多国家的粮食安全状况。FEWS NET每年多次报告粮食安全现状评估和未来预测的得分。这些得分通过五级指数量表衡量，并针对生计区的地理单位进行评估。这些区域大小各异且并非固定不变，这使得在国家内部和国家之间以及不同时间对粮食安全进行比较变得复杂。为了便于此类分析以及与其他来源的数据实现互操作性，我们将可用的原始数据转换为地理空间网格单元，其具有统一的、静态分辨率为0.5°×0.5°，这是跨不同学科研究中常用的数据格式。FEWS NET提供公开在线访问反映2009年以来报告的shapefile文件。单独的shapefile文件捕获评估和预测，并按指数得分进一步细分。每个shapefile文件可能包含一个复杂的（多）多边形，生计区之间没有明显区分。叠加地理空间网格可将（多）多边形分解为标准单元。我们对2009年7月至2020年10月期间FEWS NET在东非、南部非洲和西非区域分组中跟踪的所有25个国家（包括也门）的shapefile文件进行了网格单元转换。对于每个报告周期，根据相应生计区的原始数据，为每个网格单元分配评估得分以及短期和中期预测得分。此外，我们计算了中期预测相对于后续评估的偏差值，该值可作为准确性验证的指标。本文提供了评估和预测得分以及偏差值的网格单元数据。此外，我们展示了时间推移动画地图，作为可视化这些指标在非洲的历史模式和趋势的工具。我们的相关研究文章使用网格单元数据评估了FEWS NET预测的准确性，包括作为人道主义援助、气候条件和暴力冲突变化的函数（Backer和Billing [1]）。研究人员同样可以使用网格单元数据对粮食安全预测进行进一步验证，并研究评估和预测与潜在驱动因素及后果之间的关系。这些数据和动画对于整个国际社会中寻求广泛了解和传播有关粮食安全预测趋势知识的利益相关者也很有价值。
